CN103810668A - Double color image blind watermarking method based on singular value decomposition - Google Patents
Double color image blind watermarking method based on singular value decomposition Download PDFInfo
- Publication number
- CN103810668A CN103810668A CN201410054647.2A CN201410054647A CN103810668A CN 103810668 A CN103810668 A CN 103810668A CN 201410054647 A CN201410054647 A CN 201410054647A CN 103810668 A CN103810668 A CN 103810668A
- Authority
- CN
- China
- Prior art keywords
- watermark
- image
- layering
- embedded
- block
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a double color image blind watermarking method based on singular value decomposition. A color image is embedded into a color host image as a digital watermark, and the embedded digital watermark has high invisible performance, and therefore a copyright protection need based on color image identification is met. According to the double color image blind watermarking method, the singular value decomposition is performed on a 4*4 pixel block, the watermark is embedded into the 4*4 pixel block by modifying values of elements in the second row of the first column and the third row of the first column of a U component of the 4*4 pixel block, and then the invisible performance of the embedded watermark is improved by using a coefficient compensation and optimization method. After the image with the watermark is attacked, the embedded watermark is extracted according to relations of the modified elements of the U component of the image, and the purpose of blind detection is achieved. The double color image blind watermarking method based on the singular value decomposition has high watermark invisibility, and is applicable to copyright protection of using the color digital image as the digital watermark.
Description
Technical field
The invention belongs to field of information security technology, relate to the copyright protection of large capacity color digital image as digital watermarking.
Background technology
Along with the fast development of digital watermark technology, people just become significant coloured image from traditional pseudo-random sequence, bianry image, gray level image for the mark of copyright protection, thereby are worth extensive concern using the color digital image of large information capacity as the copyright protection technology of watermark; But, on the watermarking algorithm majority having proposed at present for gray level image, to Color Images Watermarking Algorithm, research will be lacked, this is mainly because gray level image is convenient to process compared with coloured image, and gray level image only contains monochrome information and containing chrominance information, embed watermark can not produce new color component therein, and while embedding the larger Watermarking for Color Images of quantity of information, watermark encoder, embedding and extraction, by there being larger difficulty, have reduced watermark invisibility.Therefore, as being embedded colored host image by high invisibility mode how, Watermarking for Color Images becomes one of problem demanding prompt solution.
The copyright protection information comprising due to color digital image is very large; can embed more easily or extract watermark with non-blind checking method; the blind watermark of right and wrong of existing a lot of Watermarking for Color Images method research; but in digital watermark technology research in recent years, blind Detecting digital watermark method becomes the main flow of digital watermark technology development gradually.Therefore, the blind extraction that realizes color watermark image that how to change the status quo is one of difficult point of present image digital watermark technology.
Summary of the invention
The object of this invention is to provide a kind of two coloured image blind watermark methods based on svd, comprise watermark embed process and leaching process, the detailed process that its watermark embeds is described below:
The first step: the pre-service of Watermarking for Color Images: by dimension-reduction treatment by three-dimensional color image digital watermarking
wbe divided into 3 two-dimentional layering watermarking images
w i ,
i=1,2,3 represent respectively tri-layers of R, G, B; Meanwhile, in order to improve the security of watermark, by layering watermark
w i carry out based on
kA i arnold conversion, and each decimal system pixel transitions is become to 8 scale-of-two watermark sequences;
Second step: select embedded block: original host image
halso be divided into 3 layering host images
h i ,
i=1,2,3 represent respectively tri-layers of R, G, B; Meanwhile, each layering host image
h i carry out 4 × 4 non-overlapped piecemeal, and utilize based on key
k i pseudo-random sequence at layering host image
h i middle selection embedded block is to embed layering watermark
w i ;
The 3rd step: according to formula (1) by the embedded block of choosing
acarry out svd, obtain its orthogonal matrix
u,
vwith singular value matrix
d;
The 4th step: according to formula (2) and (3), revise
umatrix first row the second row element
u 5with first row the third line element
u 9value embed scale-of-two watermark information
w, obtain the orthogonal matrix after embed watermark
;
Wherein, |
x| represent
xabsolute value,
tfor watermark embeds threshold value, sign (
x) represent
xnumerical symbol and
;
The 5th step: establish
u 5,
u 9index word be respectively
,
,
,
; Revise according to formula (4)
vthe value of middle first row element obtains compensation matrix
;
The 6th step: establish modification
u 5,
u 9after embedded block be
, the embedded block after compensation is
,
,
; Carry out the optimization method shown in formula (5), obtain the finally embedded block containing watermark
;
Wherein,
mthe sequence number of element in embedded block;
The 7th step: repeated execution of steps the 3rd step to the six steps, until all watermark informations have all embedded; Finally, layered image R, G, B containing watermark are reconfigured and obtain the image containing watermark
;
The detailed process of watermark extracting of the present invention is as follows:
The first step: the pre-service containing watermarking images: will be containing watermarking images
be divided into 3 layerings containing watermarking images
,
i=1,2,3 represent respectively tri-layers of R, G, B, and each layering is contained to watermarking images
further be divided into 4 × 4 non-overlapped block of pixels;
Second step: choose the image block containing watermark: utilize based on key
k i pseudo-random sequence at layered image
middle selection embedded block is to extract embedded layering watermark
;
The 3rd step: the image block containing watermark is carried out to svd, and obtain its orthogonal matrix
;
The 4th step: according to formula (6), utilize matrix
first row element
with
relation extract watermark
;
The 5th step: repeat above-mentioned steps second step to the four steps, until all block of pixels containing watermark are all processed, every proposed watermark information 8 are one group and are converted to decimal system numerical value; Utilize based on key
kA i (
i=1,2,3) contrary Arnold conversion converts each layering watermark, and they are reconfigured and form final watermark of extracting
.
The method simple and fast, has good watermark invisibility, can also " have no to perceive " using the color digital image of high information quantity as digital watermarking to be embedded in host image; Extract the help that does not need original host image or original watermark image when watermark, can the embedded watermark of rapid extraction from various under fire images, this invention is applicable to the copyright protection of color digital image as digital watermarking.
Accompanying drawing explanation
Fig. 1 (a), Fig. 1 (b) are two width original color host images.
Fig. 2 is a width color watermark image.
Fig. 3 (a), Fig. 3 (b) be by the watermark shown in Fig. 2 be embedded into successively obtain after host image Fig. 1 (a), Fig. 1 (b) containing watermarking images, its structural similarity SSIM value is 0.97646,0.93304 successively, and its Y-PSNR PSNR value is 33.8332dB, 35.6479dB successively.
Fig. 4 (a), Fig. 4 (b) are the watermarks of extracting from Fig. 3 (a), Fig. 3 (b) successively, and its normalized crosscorrelation coefficient NC value is respectively 1.00000,1.00000.
Fig. 5 (a), Fig. 5 (b), Fig. 5 (c), Fig. 5 (d), Fig. 5 (e), Fig. 5 (f) be by shown in Fig. 3 (a) containing watermarking images carry out successively JPEG compression, add make an uproar, watermark that attacks such as low-pass filtering, sharpening, convergent-divergent, shearing extracted afterwards, its normalized crosscorrelation coefficient NC value is respectively 0.99611,0.95336,0.89987,0.99994,0.94319,0.83351.
Embodiment
The object of this invention is to provide a kind of two coloured image blind watermark methods based on svd, comprise concrete watermark embed process and leaching process, describe its watermark embed process by example as follows:
The first step: the pre-service of Watermarking for Color Images: by dimension-reduction treatment by 24 three-dimensional color image digital watermarkings of 32 × 32
wbe divided into 3 two-dimentional layering watermarking images
w i ,
i=1,2,3 represent respectively tri-layers of R, G, B; In order to improve the security of watermark, by each layering watermark carry out based on
kA i arnold conversion, and each decimal system pixel transitions is become to 8 scale-of-two watermark sequences, for example can be by 246,68,7 convert respectively binary number 11110110,01000100,00000111 to;
Second step: select embedded block: by 24 original host images of true color of 512 × 512
halso be divided into 3 layering host images
h i ,
i=1,2,3 represent respectively tri-layers of R, G, B; Meanwhile, each layering host image is carried out to 4 × 4 non-overlapped piecemeal, and utilize based on key
k i pseudo-random sequence at layering host image
h i middle selection embedded block is to embed layering watermark
w i ;
The 3rd step: according to formula (1) by the embedded block of choosing
a=
carry out svd, obtain its orthogonal matrix
u= ,
The 4th step: according to formula (2) and (3), by revising
umatrix first row the second row element
u 5with first row the third line element
u 9value carry out embed watermark information
, establish watermark and embed threshold value
t=0.0400, obtain embedded matrix
=
;
if (3)
Wherein, |
x| represent
xabsolute value,
tfor watermark embeds threshold value, sign (
x) represent
xnumerical symbol and
;
The 5th step: establish
u 5,
u 9index word be respectively
,
,
=0.0201,
=-0.0201, revise according to formula (4)
vthe value of middle first row element obtains compensation matrix
=
;
Wherein,
;
According to formula (5), carry out optimization method, obtain finally containing the embedded block of watermark be
=
;
Wherein,
mthe sequence number of element in embedded block;
The 7th step: repeated execution of steps the 3rd step to the six steps, until all watermark informations are all embedded into the block of pixels of selection; Finally, layered image R, G, B containing watermark are reconfigured and obtain the image containing watermark
;
The detailed process of watermark extracting of the present invention is as follows:
The first step: the pre-service containing watermarking images: will be containing watermarking images
be divided into 3 layerings containing watermarking images
,
i=1,2,3 represent respectively tri-layers of R, G, B, and each layering is contained to watermarking images
further be divided into 4 × 4 non-overlapped block of pixels;
Second step: choose the image block containing watermark: utilize based on key
k i pseudo-random sequence at layered image
middle selection embedded block is to extract embedded layering watermark
;
The 3rd step: to the image block containing watermark
carry out svd and obtain its orthogonal matrix
=
;
The 4th step: according to formula (6), utilize matrix
first row element
with
relation extract watermark
;
(6)
The 5th step: repeat above-mentioned steps second step to the four steps, until all block of pixels containing watermark are all processed, every proposed watermark information 8 are one group and are converted to decimal system numerical value; Utilize based on key
kA i contrary Arnold conversion each layering watermark is converted, and they are reconfigured and form final watermark of extracting
.
The method simple and fast, has higher watermark invisibility and robustness, is applicable to the copyright protection of coloured image as digital watermarking.
validation verification of the present invention
In order to prove validity of the present invention, 24 standard pictures that the two width sizes of selection as shown in Fig. 1 (a), Fig. 1 (b) are 512 × 512 are as host image, and 24 coloured images that are 32 × 32 using a width size are as shown in Figure 2 verified as digital watermarking.
Fig. 3 (a), Fig. 3 (b) be by the watermark shown in Fig. 2 be embedded into successively obtain after host image Fig. 1 (a), Fig. 1 (b) containing watermarking images, its structural similarity SSIM value is 0.97646,0.93304 successively, its Y-PSNR PSNR value is 33.8332dB, 35.6479dB successively, can find out that embedded watermark has good invisibility;
Fig. 4 (a), Fig. 4 (b) are the watermarks of extracting from Fig. 3 (a), Fig. 3 (b) successively, and its normalized crosscorrelation coefficient NC value is respectively 1.00000,1.00000, Fig. 5 (a), Fig. 5 (b), Fig. 5 (c), Fig. 5 (d), Fig. 5 (e), the watermarking images that contains shown in Fig. 3 (a) is carried out successively JPEG compression by Fig. 5 (f), add and make an uproar, low-pass filtering, sharpening, convergent-divergent, the watermark of extracting after the attacks such as shearing, its normalized crosscorrelation coefficient NC value is respectively 0.99611, 0.95336, 0.89987, 0.99994, 0.94319, 0.83351, as can be seen here, the digital watermarking image extracting has good identifiability, illustrate that the method has stronger robustness, can extract well embedded color watermark.
Claims (1)
1. the two coloured image blind watermark methods based on svd, is characterized in that: comprise watermark embed process and leaching process, the detailed process that its watermark embeds is described below:
The first step: the pre-service of Watermarking for Color Images: by dimension-reduction treatment by three-dimensional color image digital watermarking
wbe divided into 3 two-dimentional layering watermarking images
w i ,
i=1,2,3 represent respectively tri-layers of R, G, B; Meanwhile, in order to improve the security of watermark, by layering watermark
w i carry out based on
kA i arnold conversion, and each decimal system pixel transitions is become to 8 scale-of-two watermark sequences;
Second step: select embedded block: original host image
halso be divided into 3 layering host images
h i ,
i=1,2,3 represent respectively tri-layers of R, G, B; Meanwhile, each layering host image
h i carry out 4 × 4 non-overlapped piecemeal, and utilize based on key
k i pseudo-random sequence at layering host image
h i middle selection embedded block is to embed layering watermark
w i ;
The 3rd step: according to formula (1) by the embedded block of choosing
acarry out svd, obtain its orthogonal matrix
u,
vwith singular value matrix
d;
The 4th step: according to formula (2) and (3), revise
umatrix first row the second row element
u 5with first row the third line element
u 9value embed scale-of-two watermark information
w, obtain the orthogonal matrix after embed watermark
;
Wherein, |
x| represent
xabsolute value,
tfor watermark embeds threshold value, sign (
x) represent
xnumerical symbol and
;
The 5th step: establish
u 5,
u 9index word be respectively
,
,
,
; Revise according to formula (4)
vthe value of middle first row element obtains compensation matrix
;
The 6th step: establish modification
u 5,
u 9after embedded block be
, the embedded block after compensation is
,
,
; Carry out the optimization method shown in formula (5), obtain the finally embedded block containing watermark
;
Wherein,
mthe sequence number of element in embedded block;
The 7th step: repeated execution of steps the 3rd step to the six steps, until all watermark informations have all embedded; Finally, layered image R, G, B containing watermark are reconfigured and obtain the image containing watermark
;
The detailed process of watermark extracting of the present invention is as follows:
The first step: the pre-service containing watermarking images: will be containing watermarking images
be divided into 3 layerings containing watermarking images
,
i=1,2,3 represent respectively tri-layers of R, G, B, and each layering is contained to watermarking images
further be divided into 4 × 4 non-overlapped block of pixels;
Second step: choose the image block containing watermark: utilize based on key
k i pseudo-random sequence at layered image
middle selection embedded block is to extract embedded layering watermark
;
The 3rd step: the image block containing watermark is carried out to svd, and obtain its orthogonal matrix
;
The 4th step: according to formula (6), utilize matrix
first row element
with
relation extract watermark
;
The 5th step: repeat above-mentioned steps second step to the four steps, until all block of pixels containing watermark are all processed, every proposed watermark information 8 are one group and are converted to decimal system numerical value; Utilize based on key
kA i (
i=1,2,3) contrary Arnold conversion converts each layering watermark, and they are reconfigured and form final watermark of extracting
.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410054647.2A CN103810668A (en) | 2014-02-19 | 2014-02-19 | Double color image blind watermarking method based on singular value decomposition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410054647.2A CN103810668A (en) | 2014-02-19 | 2014-02-19 | Double color image blind watermarking method based on singular value decomposition |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103810668A true CN103810668A (en) | 2014-05-21 |
Family
ID=50707391
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410054647.2A Pending CN103810668A (en) | 2014-02-19 | 2014-02-19 | Double color image blind watermarking method based on singular value decomposition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103810668A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106157233A (en) * | 2016-07-11 | 2016-11-23 | 鲁东大学 | A kind of good digital water mark method decomposed based on matrix Schur |
CN106169171A (en) * | 2016-07-15 | 2016-11-30 | 鲁东大学 | The good digital water mark method decomposed based on Hessenberg |
CN107274333A (en) * | 2017-06-21 | 2017-10-20 | 鲁东大学 | A kind of spatial domain coloured image blind watermark method for merging singular value decomposition |
CN108460716A (en) * | 2018-01-30 | 2018-08-28 | 大连工业大学 | Spot color printing image digital watermark method and system |
CN105427231B (en) * | 2015-11-26 | 2018-11-06 | 南京林业大学 | A kind of SVD double-layer digital water mark methods avoiding false alarm |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101866477A (en) * | 2010-06-11 | 2010-10-20 | 山东大学 | Digital blind watermarking method for double protection of image integrity |
CN102496367A (en) * | 2011-12-06 | 2012-06-13 | 北京邮电大学 | Audio blind watermark algorithm based on combination of discrete wavelet transformation (DWT), discrete cosine transformation (DCT) and singular value decomposition (SVD) |
-
2014
- 2014-02-19 CN CN201410054647.2A patent/CN103810668A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101866477A (en) * | 2010-06-11 | 2010-10-20 | 山东大学 | Digital blind watermarking method for double protection of image integrity |
CN102496367A (en) * | 2011-12-06 | 2012-06-13 | 北京邮电大学 | Audio blind watermark algorithm based on combination of discrete wavelet transformation (DWT), discrete cosine transformation (DCT) and singular value decomposition (SVD) |
Non-Patent Citations (3)
Title |
---|
M. Q. FAN等: "Restudy on SVD-based watermarking scheme", 《APPLIED MATHEMATICS AND COMPUTATION》 * |
苏庆堂: "基于盲提取的彩色图像数字水印算法", 《中国博士学位论文全文数据库 信息科技辑》 * |
范明泉: "数字图像水印与量化音频水印的若干关键技术研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105427231B (en) * | 2015-11-26 | 2018-11-06 | 南京林业大学 | A kind of SVD double-layer digital water mark methods avoiding false alarm |
CN106157233A (en) * | 2016-07-11 | 2016-11-23 | 鲁东大学 | A kind of good digital water mark method decomposed based on matrix Schur |
CN106157233B (en) * | 2016-07-11 | 2019-11-05 | 鲁东大学 | A kind of good digital water mark method decomposed based on matrix Schur |
CN106169171A (en) * | 2016-07-15 | 2016-11-30 | 鲁东大学 | The good digital water mark method decomposed based on Hessenberg |
CN106169171B (en) * | 2016-07-15 | 2019-08-13 | 鲁东大学 | The good digital water mark method decomposed based on Hessenberg |
CN107274333A (en) * | 2017-06-21 | 2017-10-20 | 鲁东大学 | A kind of spatial domain coloured image blind watermark method for merging singular value decomposition |
CN107274333B (en) * | 2017-06-21 | 2020-05-05 | 鲁东大学 | Spatial domain color image blind watermarking method fusing singular value decomposition |
CN108460716A (en) * | 2018-01-30 | 2018-08-28 | 大连工业大学 | Spot color printing image digital watermark method and system |
CN108460716B (en) * | 2018-01-30 | 2022-05-10 | 大连工业大学 | Digital watermarking method and system for spot color printing image |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Emad et al. | A secure image steganography algorithm based on least significant bit and integer wavelet transform | |
CN103810668A (en) | Double color image blind watermarking method based on singular value decomposition | |
CN104050622A (en) | Color image blind-watermarking method based on ternary coding | |
CN107895340B (en) | Space-frequency domain color digital image blind watermarking method integrating QR decomposition | |
CN104899823A (en) | Two-color image blind watermarking method based on Hessenberg disintegration | |
CN108648134B (en) | Spatial domain color digital image blind watermarking method fusing discrete wavelet transform | |
CN104168484A (en) | Video watermarking method based on visual attention mechanism | |
CN110796586B (en) | Blind watermarking method and system based on digital dot matrix and readable storage medium | |
CN109035129A (en) | A kind of color digital image blind watermark method based on two-dimensional discrete sine transform | |
CN102663669B (en) | Dual-color digital image blind watermarking method based on Schur decomposition | |
CN103617591B (en) | A kind of dual-color image blind watermarking method decomposed based on QR and compensate | |
CN101916427B (en) | Spatial domain-based image watermark adding method | |
CN104537601A (en) | Gray level image aerial region steganography method based on nine grids | |
CN103150701A (en) | QR (Quick Response) code watermarking algorithm with strong robustness | |
CN102523440B (en) | Blind digital watermarking encoding and decoding method capable of withstanding cropping attack for fusion vision model | |
CN107274333A (en) | A kind of spatial domain coloured image blind watermark method for merging singular value decomposition | |
CN105023236B (en) | A kind of dual-color image blind watermarking method based on LU Decomposition | |
CN103778592A (en) | Double-color image blind watermarking method based on ary excess code | |
CN104063832B (en) | A kind of watermark insertion and extracting method and device | |
CN103310130B (en) | Text document embedding algorithm and extracting method | |
CN106991636B (en) | Airspace color image blind watermarking method fusing approximate Schur decomposition | |
CN103024375A (en) | Three-dimensional image semi-fragile watermarking method based on binocular just-perceived distortion | |
CN102156956B (en) | High robustness watermark method based on singular value decomposition and discrete cosine transform | |
CN106157233B (en) | A kind of good digital water mark method decomposed based on matrix Schur | |
CN110417551A (en) | A kind of the character picture production camouflage and restoration methods of the diffusion of combination error |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140521 |